Pakistan joined the IMF in 1950. Since then it has entered 25 formal lending arrangements, plus a separate emergency Rapid Financing Instrument in 2020. That works out to one arrangement every three years. The 1990s alone saw six, and the 2020s have already produced three.
Those numbers are the place to start any honest assessment of where Pakistan stands today. Each programme has been necessary at the time it was signed, and several prevented disorderly default. But 25 arrangements have not produced an economy that can stay outside IMF support for any sustained period. A borrower that returns this often is not having bad luck. Its frequent returns are an institutional outcome.
The IMF has said as much itself. In 2005 it described Pakistan as one of its most prolonged users of Fund resources, having spent roughly half of the previous 46 years under IMF-supported programmes. The same retrospective called the 2000–04 programmes a success: they were completed, growth accelerated, inflation was low and the external position improved. It also conceded that progress had been slower wherever institution-building was required. Pakistan was back in a programme in 2008.
That history should make everyone cautious about declaring victory in 2026.
What has actually been stabilized
The recent gains are real. Reserves have recovered substantially, the primary balance has improved, inflation is far below its crisis peak, the exchange market functions better and near-term financing has been secured. The IMF fairly describes this as progress in stabilizing the economy and rebuilding confidence.
But stabilizing indicators is not the same as stabilizing an economy. Several measures have barely moved:
• Investment was 14.38 percent of GDP in FY2026, roughly where it has been for years. The Economic Survey itself concedes this is low by regional standards.
• Private fixed investment remains below 10 percent of GDP.
• Exports remain below 10 percent of GDP.
• Medium-term growth, in the IMF’s own projections, settles around 4–4.5 percent. That is a recovery path, not an Asian-style transformation.
If confidence has truly returned, the exchange rate is more credible and interest rates are falling from their peak, why has investment not responded? The likely answer is that an investor deciding on a ten-year factory does not care that the current account balances this quarter. She needs confidence about Pakistan’s rules over the next decade, and that credibility has not been built.
So the real test of stabilization is not whether reserves have risen. It is whether Pakistan can grow at 5–7 percent without generating another balance-of-payments crisis. There is no evidence yet that it can.
Reserves are a stock; financing is a flow
The reserves story, taken alone, misleads. Reserves are a stock. Financing requirements are a recurring flow.
IMF projections put Pakistan’s gross external financing requirements at about USD 122.7 billion over FY2027–FY2031: USD 21.2bn in FY27, USD 29.9bn in FY28, USD 23.6bn in FY29, USD 22.1bn in FY30 and USD 26.0bn in FY31. A country can hold USD 20–25 billion in reserves and remain fragile if it must raise or roll over a comparable sum every year.
The narrower programme financing gap, USD 11.6 billion over FY25–28 excluding the Resilience and Sustainability Facility, understates the problem. The IMF explicitly excludes from that figure the rollover of short-term bilateral deposits and expected commercial-bank refinancing. The USD 11.6 billion is what remains after all of that is assumed to go smoothly.
The relevant question is therefore not “do we have enough reserves?” It is how much foreign exchange the economy earns, how much must be rolled over, and how much still depends on the IMF, friendly capitals, multilaterals and commercial lenders. That is the difference between borrowed liquidity and earned resilience.
Sustainable is not the same as comfortable
The IMF judges Pakistan’s public debt sustainable, but with high medium-term risks and a narrow path. Its baseline has general government debt falling from 72.8 percent of GDP in FY25 to 58.9 percent by FY31. That path assumes several things hold at once: persistent primary surpluses, steady nominal growth, continued bilateral rollovers, no major exchange-rate shock, and containment of SOE and energy liabilities. In other words, it assumes the old boom-bust cycle does not return.
Pakistan actually faces three separate debt tests:
Fiscal solvency: can revenues service public debt?
Liquidity: can maturing debt be refinanced?
External repayment capacity: can the economy earn enough foreign exchange without compressing imports or arranging exceptional financing?
The IMF baseline says the first is manageable. The financing wall shows the second remains difficult. Exports below 10 percent of GDP cast doubt on the third. Debt-to-GDP measures solvency against national income. It says little about the foreign-exchange constraint, which is the one that has repeatedly brought Pakistan down.
The budget hides a quasi-fiscal state
The headline fiscal deficit understates the state’s true footprint. In early 2026, power-sector circular debt stood at Rs1.764 trillion and gas circular debt at Rs3.442 trillion, together Rs5.2 trillion. Federal loss-making SOEs lost Rs832.8 billion in FY2025, while government fiscal support to SOEs reached Rs2.079 trillion. SOE debt is Rs9.57 trillion, sovereign guarantees Rs2.164 trillion, and unfunded SOE pension liabilities roughly Rs2 trillion.
Some of these figures overlap, so they cannot simply be added together. That is precisely the point: no single budget line captures the public-sector balance sheet. The Finance Division’s own Fiscal Risk Statement estimates that absorbing circular debt into public debt could raise the debt ratio by about 4 percentage points and add roughly Rs937 billion a year in debt service over FY2026–28.
The real fiscal deficit should be measured as the budget deficit, plus SOE losses, plus new circular debt, plus crystallizing guarantees, plus unfunded commitments. Pakistan also holds vast underperforming public assets, including urban land, SOEs, buildings and mineral rights. Debt management should become public wealth management.
Energy: how Pakistan does reform badly
Energy is the clearest case of reform failing. Governments, the IMF, the World Bank and the ADB have all worked on it for decades, yet circular debt still exceeds Rs5.2 trillion. The World Bank approved yet another USD 375.9 million grid project in July 2026. PIDE’s 2022 volume Power Sector: An Enigma With No Easy Solution, co-edited by Nadeem ul Haque, had already described the contradiction: Pakistan moved from shortage to excess capacity while circular debt, cash-flow crises and borrowing to cover sector deficits persisted.
The core error is treating the problem as tariffs, subsidies and generation capacity when it is really about market structure, contracts and incentives. Cost recovery is not cost reduction. Raising tariffs to recover capacity payments, theft, collection failures and bad legacy contracts can balance an accounting identity while destroying industrial demand. The consumer becomes the residual payer for failures elsewhere in the system. As Haque and Shahid Sattar have argued, Pakistan cannot tax its future to pay for its energy past.
Circular debt is simply a delayed fiscal deficit. Claiming that its flow has been “contained” is insufficient while the sector still cannot survive without sovereign guarantees. What Pakistan needs is genuine open access, a competitive wholesale market, direct contracting, independent market operation, commercially disciplined DISCOs, and legacy debt separated from current tariffs.
Institutions, not ribbons
The 2024 Economics Nobel rewarded work showing that institutions explain persistent differences in prosperity. Pakistan’s recent reforms deserve acknowledgement: the Asaan Karobar Act, a planned Regulatory Quality Control Office, a regulatory registry and some tariff liberalization. But these are commitments and mechanisms. They are not yet evidence that the institutional equilibrium has changed.
A registry that publishes 100 permissions solves little if 70 of them should not exist. The question to ask is why a permission exists, not whether it can be digitized. Digitized sludge is still sludge.
Meanwhile, the governing machinery still runs on the cycle of PC-1, PSDP allocation, procurement and inauguration. That cycle serves ministries, politicians, contractors, consultants and donors who need to disburse. Abolishing an agency or giving power to cities offers no ribbon to cut. Projects are not reform.
The real test
Pakistan has controlled an acute crisis, yet again. That is the first stage. Durable stabilization, the second stage, means ending dependence on exceptional financing and eliminating recurrent quasi-fiscal deficits. Transformation, the third, means investment rising toward 20–25 percent of GDP and exports growing faster than GDP. Neither has yet begun.
The test of the current programme is not whether Pakistan completes it. Pakistan has completed programmes before. The test is whether the economy, the energy system, the state and the financing model are different enough at the end that a 26th arrangement is unnecessary.
Copyright Business Recorder, 2026
The writer served as the Deputy Chairman of the Planning Commission. X: @nadeemhaque; YouTube: @SiaLytics and Substack: Aid, Policy and Growth